Enable subgraph metrics and extended error metrics by default

.changesets/breaking_zach_router_1910_telemetry_defaults.md

@@ -0,0 +1,24 @@
+### Enable subgraph metrics and extended error metrics by default
+
+Router 3.0 changes the defaults for two Apollo telemetry settings to match the preferred configuration for GraphOS Studio users, and promotes one of them out of preview.
+
+**`telemetry.apollo.subgraph_metrics` now defaults to `true`.** Subgraph metrics send additional per-subgraph operation metrics to GraphOS Studio via OTLP, powering subgraph insights. Previously this was opt-in. To restore the previous behavior, set:
+
+```yaml
+telemetry:
+  apollo:
+    subgraph_metrics: false
+```
+
+**`telemetry.apollo.errors.preview_extended_error_metrics` has been renamed to `telemetry.apollo.errors.extended_error_metrics` and now defaults to `enabled`.** Extended error metrics send OTLP error metrics with additional dimensions (`extensions.service`, `extensions.code`), giving Studio richer error attribution out of the box. The `preview_` prefix has been dropped now that the feature is stable.
+
+Configurations using the old `preview_extended_error_metrics` field name are migrated automatically at startup (with a warning). To restore the previous behavior, set:
+
+```yaml
+telemetry:
+  apollo:
+    errors:
+      extended_error_metrics: disabled
+```
+
+By [@BobaFetters](https://github.com/BobaFetters) in https://github.com/apollographql/router/pull/9879

apollo-router/src/configuration/metrics.rs

@@ -409,8 +409,8 @@ impl InstrumentData {
             "$.signature_normalization_algorithm",
             opt.metrics_reference_mode,
             "$.metrics_reference_mode",
-            opt.errors.preview_extended_error_metrics,
-            "$.errors.preview_extended_error_metrics",
+            opt.errors.extended_error_metrics,
+            "$.errors.extended_error_metrics",
             opt.field_level_instrumentation_sampler,
             "$.field_level_instrumentation_sampler",
             opt.tracing.batch_processor.scheduled_delay,

apollo-router/src/configuration/migrations/3003-extended-error-metrics.yaml

@@ -0,0 +1,5 @@
+description: preview_extended_error_metrics is now extended_error_metrics
+actions:
+  - type: move
+    from: telemetry.apollo.errors.preview_extended_error_metrics
+    to: telemetry.apollo.errors.extended_error_metrics

apollo-router/src/configuration/snapshots/apollo_router__configuration__metrics__test__metrics@apollo_telemetry.router.yaml.snap

@@ -7,7 +7,7 @@ expression: "& metrics.non_zero()"
     datapoints:
       - value: 1
         attributes:
-          opt.errors.preview_extended_error_metrics: enabled
+          opt.errors.extended_error_metrics: enabled
           opt.field_level_instrumentation_sampler: 1
           opt.metrics.otlp.batch_processor.max_export_timeout: 103s
           opt.metrics.otlp.batch_processor.scheduled_delay: 5s

apollo-router/src/configuration/snapshots/apollo_router__configuration__metrics__test__metrics@experimental_features.router.yaml.snap

@@ -7,7 +7,7 @@ expression: "& metrics.non_zero()"
     datapoints:
       - value: 1
         attributes:
-          opt.errors.preview_extended_error_metrics: false
+          opt.errors.extended_error_metrics: false
           opt.field_level_instrumentation_sampler: false
           opt.metrics.otlp.batch_processor.max_export_timeout: false
           opt.metrics.otlp.batch_processor.scheduled_delay: false

apollo-router/src/configuration/snapshots/apollo_router__configuration__tests__schema_generation.snap

@@ -305,7 +305,7 @@ expression: "&schema"
           "description": "Set the signature normalization algorithm to use when sending Apollo usage reports."
         },
         "subgraph_metrics": {
-          "default": false,
+          "default": true,
           "description": "Enable sending additional subgraph metrics to Apollo Studio via OTLP",
           "type": "boolean"
         },
@@ -3652,7 +3652,7 @@ expression: "&schema"
     "ErrorsConfiguration": {
       "additionalProperties": false,
       "properties": {
-        "preview_extended_error_metrics": {
+        "extended_error_metrics": {
           "allOf": [
             {
               "$ref": "#/definitions/ExtendedErrorMetricsMode"

apollo-router/src/configuration/snapshots/apollo_router__configuration__tests__upgrade_old_configuration@apollo_extended_error_metrics.yaml.snap

@@ -0,0 +1,9 @@
+---
+source: apollo-router/src/configuration/tests.rs
+expression: new_config
+---
+---
+telemetry:
+  apollo:
+    errors:
+      extended_error_metrics: enabled

apollo-router/src/configuration/snapshots/apollo_router__configuration__tests__upgrade_old_configuration@apollo_extended_errors.yaml.snap

@@ -6,4 +6,4 @@ expression: new_config
 telemetry:
   apollo:
     errors:
-      preview_extended_error_metrics: enabled
+      extended_error_metrics: enabled

apollo-router/src/configuration/testdata/metrics/apollo_telemetry.router.yaml

@@ -4,7 +4,7 @@ supergraph:
 telemetry:
   apollo:
     errors:
-      preview_extended_error_metrics: enabled
+      extended_error_metrics: enabled
       subgraph:
         all:
           send: true

apollo-router/src/configuration/testdata/migrations/apollo_extended_error_metrics.yaml

@@ -0,0 +1,4 @@
+telemetry:
+  apollo:
+    errors:
+      preview_extended_error_metrics: enabled

apollo-router/src/plugins/connectors/incompatible/telemetry.rs

@@ -23,7 +23,7 @@ impl IncompatiblePlugin for TelemetryIncompatPlugin {
         self.config.subgraph.all.send
             // When ExtendedErrorMetricsMode is enabled, this plugin supports reporting connector errors
             && !matches!(
-                self.config.preview_extended_error_metrics,
+                self.config.extended_error_metrics,
                 apollo::ExtendedErrorMetricsMode::Enabled
             )
     }
@@ -39,7 +39,7 @@ impl IncompatiblePlugin for TelemetryIncompatPlugin {
                 .partition_map(|(name, sub)| {
                     if sub.send
                         && !matches!(
-                            self.config.preview_extended_error_metrics,
+                            self.config.extended_error_metrics,
                             apollo::ExtendedErrorMetricsMode::Enabled
                         )
                     {
@@ -61,13 +61,13 @@ impl IncompatiblePlugin for TelemetryIncompatPlugin {
             if self.config.subgraph.subgraphs.contains_key(subgraph) {
                 tracing::warn!(
                     subgraph = subgraph,
-                    message = "plugin `telemetry` is explicitly configured to send errors to Apollo studio for connector-enabled subgraph, which is only supported when `preview_extended_error_metrics` is enabled",
+                    message = "plugin `telemetry` is explicitly configured to send errors to Apollo studio for connector-enabled subgraph, which is only supported when `extended_error_metrics` is enabled",
                     see = "https://go.apollo.dev/connectors/incompat",
                 );
             } else {
                 tracing::info!(
                     subgraph = subgraph,
-                    message = "plugin `telemetry` is indirectly configured to send errors to Apollo studio for a connector-enabled subgraph, which is only supported when `preview_extended_error_metrics` is enabled",
+                    message = "plugin `telemetry` is indirectly configured to send errors to Apollo studio for a connector-enabled subgraph, which is only supported when `extended_error_metrics` is enabled",
                     see = "https://go.apollo.dev/connectors/incompat",
                 );
             }

apollo-router/src/plugins/telemetry/apollo.rs

@@ -277,7 +277,7 @@ pub(crate) struct ErrorsConfiguration {
     pub(crate) subgraph: SubgraphErrorConfig,
 
     /// Send error metrics via OTLP with additional dimensions [`extensions.service`, `extensions.code`]
-    pub(crate) preview_extended_error_metrics: ExtendedErrorMetricsMode,
+    pub(crate) extended_error_metrics: ExtendedErrorMetricsMode,
 }
 
 #[derive(Debug, Clone, Deserialize, JsonSchema, Default, PartialEq)]
@@ -326,10 +326,10 @@ impl SubgraphErrorConfig {
 #[serde(deny_unknown_fields, rename_all = "lowercase")]
 pub(crate) enum ExtendedErrorMetricsMode {
     /// Do not send extended OTLP error metrics
-    #[default]
     Disabled,
     /// Send extended OTLP error metrics to Apollo Studio with additional dimensions [`extensions.service`, `extensions.code`].
     /// If enabled, it's also recommended to enable `redaction_policy: extended` on subgraphs to send the `extensions.code` for subgraph errors.
+    #[default]
     Enabled,
 }
 
@@ -416,7 +416,7 @@ impl Default for Config {
             signature_normalization_algorithm: ApolloSignatureNormalizationAlgorithm::default(),
             experimental_local_field_metrics: false,
             metrics_reference_mode: ApolloMetricsReferenceMode::default(),
-            subgraph_metrics: false,
+            subgraph_metrics: true,
             sampler: None,
         }
     }

apollo-router/src/plugins/telemetry/config_new/fixtures/connector/custom_counter_with_conditions/metrics.snap

@@ -25,6 +25,17 @@ info:
                     - user_api
                     - connector_source: name
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 2
+        attributes:
+          connector.source: user_api
+          has_errors: false
+          subgraph.name: users
 - name: not.found.count
   description: Count of 404 responses from the user API
   unit: count

apollo-router/src/plugins/telemetry/config_new/fixtures/connector/custom_histogram/metrics.snap

@@ -23,6 +23,17 @@ info:
                     - user_api
                     - connector_source: name
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 2
+        attributes:
+          connector.source: user_api
+          has_errors: false
+          subgraph.name: users
 - name: rate.limit
   description: Rate limit remaining
   unit: count

apollo-router/src/plugins/telemetry/config_new/fixtures/connector/http_client_request_duration/metrics.snap

@@ -30,9 +30,21 @@ info:
               custom.supergraph.operation.kind:
                 supergraph_operation_kind: string
               custom.has_error:
-                error: boolean
-snapshot_kind: text
+                connector_on_response_error: true
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 1
+        attributes:
+          connector.source: posts_api
+          graphql.operation.name: Test
+          graphql.operation.type: query
+          has_errors: false
+          subgraph.name: posts
 - name: http.client.request.duration
   description: Duration of HTTP client requests.
   unit: s

apollo-router/src/plugins/telemetry/config_new/fixtures/connector/mapping_problems/metrics.snap

@@ -30,8 +30,18 @@ info:
                 connector_response_mapping_problems: boolean
             unit: count
             type: counter
-snapshot_kind: text
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 2
+        attributes:
+          connector.source: user_api
+          has_errors: false
+          subgraph.name: users
 - name: request.mapping.problems
   description: Count of connectors request mapping problems
   unit: count

apollo-router/src/plugins/telemetry/config_new/fixtures/connector/subgraph_and_connector/metrics.snap

@@ -16,6 +16,22 @@ info:
             attributes:
               subgraph.name: true
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 1
+        attributes:
+          connector.source: reviews_api
+          has_errors: false
+          subgraph.name: reviews
+      - sum: 0.1
+        count: 1
+        attributes:
+          has_errors: false
+          subgraph.name: products
 - name: http.client.request.duration
   description: Duration of HTTP client requests.
   unit: s

apollo-router/src/plugins/telemetry/config_new/fixtures/subgraph/caching/metrics.snap

@@ -75,6 +75,17 @@ info:
               supergraph.operation.name:
                 supergraph_operation_name: string
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 1
+        attributes:
+          graphql.operation.name: Test
+          has_errors: false
+          subgraph.name: products
 - name: apollo.router.response.cache
   description: Response cache hit/miss operations at the subgraph level
   unit: ops

apollo-router/src/plugins/telemetry/config_new/fixtures/subgraph/custom_counter/metrics.snap

@@ -1,20 +1,29 @@
 ---
 source: apollo-router/src/plugins/telemetry/config_new/instruments.rs
-description: Test server request body size metrics
+description: Custom counter
 expression: "&metrics.all()"
 info:
   telemetry:
     instrumentation:
       instruments:
-        router:
-          http.server.active_requests: false
-          http.server.request.duration: false
+        default_requirement_level: none
+        subgraph:
           custom_counter:
             description: count of requests
             type: counter
             unit: unit
             value: unit
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 1
+        attributes:
+          has_errors: false
+          subgraph.name: products
 - name: custom_counter
   description: count of requests
   unit: unit

apollo-router/src/plugins/telemetry/config_new/fixtures/subgraph/custom_counter_aborted_request/metrics.snap

@@ -1,20 +1,29 @@
 ---
 source: apollo-router/src/plugins/telemetry/config_new/instruments.rs
-description: Test server request body size metrics
+description: "Custom counter aborted request, the supergraph response didn't happen, but request should increment the metric on Drop."
 expression: "&metrics.all()"
 info:
   telemetry:
     instrumentation:
       instruments:
-        router:
-          http.server.active_requests: false
-          http.server.request.duration: false
+        default_requirement_level: none
+        subgraph:
           custom_counter:
             description: count of requests
             type: counter
             unit: unit
             value: unit
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 1
+        attributes:
+          has_errors: false
+          subgraph.name: products
 - name: custom_counter
   description: count of requests
   unit: unit

apollo-router/src/plugins/telemetry/config_new/fixtures/subgraph/custom_counter_custom_value/metrics.snap

@@ -1,21 +1,30 @@
 ---
 source: apollo-router/src/plugins/telemetry/config_new/instruments.rs
-description: Test server request body size metrics
+description: Custom counter that gets a value from a header
 expression: "&metrics.all()"
 info:
   telemetry:
     instrumentation:
       instruments:
-        router:
-          http.server.active_requests: false
-          http.server.request.duration: false
+        default_requirement_level: none
+        subgraph:
           custom_counter:
             description: count of requests
             type: counter
             unit: unit
             value:
-              request_header: count_header
+              subgraph_request_header: count_header
 ---
+- name: apollo.router.operations.fetch.duration
+  description: Duration of a subgraph fetch.
+  unit: s
+  data:
+    datapoints:
+      - sum: 0.1
+        count: 1
+        attributes:
+          has_errors: false
+          subgraph.name: products
 - name: custom_counter
   description: count of requests
   unit: unit